Python: 使用 pandas 將長格式數據轉換為寬格式(樞紐分析)

import pandas as pd
import numpy as np

# 創建示例數據
data = {
    'Net': ['M_CPU0A_SA_DQS', 'M_CPU0A_SA_DQS', 'M_CPU0A_SA_DQ', 'M_CPU0A_SA_DQ'],
    'Rank': ['N0.C0.D0.SC0.R0', 'N0.C0.D0.SC0.R0', 'N0.C0.D0.SC0.R0', 'N0.C0.D0.SC0.R0'],
    'Test': ['RxDqs-', 'RxDqs+', 'RxDqs-', 'RxDqs+'],
    'Mean': [125.3, 118.7, 135.2, 127.9],
    'Std Dev': [3.2, 2.9, 3.5, 3.1]
}

df = pd.DataFrame(data)
print("原始長格式數據：")
print(df)

# 檢查是否有重複的 (Net, Rank, Test) 組合
duplicates = df.duplicated(subset=['Net', 'Rank', 'Test'], keep=False)
if duplicates.any():
    print("警告：數據中存在重複的 (Net, Rank, Test) 組合！")
    print(df[duplicates])
else:
    print("數據中沒有重複的 (Net, Rank, Test) 組合。")

# 初始化一個空的 DataFrame 用於存儲寬格式結果
df_wide = pd.DataFrame()

# 為每個統計量創建一個透視表
for value_col in ['Mean', 'Std Dev']:
    # 使用 pivot 函數將 Test 轉為列
    pivot = df.pivot(index=['Net', 'Rank'], columns='Test', values=value_col)
    
    # 重命名列以包含統計量
    pivot.columns = [f"{test} {value_col}" for test in pivot.columns]
    
    # 如果是第一個統計量，初始化 df_wide
    if df_wide.empty:
        df_wide = pivot
    else:
        # 否則，使用 join 將新的透視表添加到 df_wide
        df_wide = df_wide.join(pivot)

# 重設索引，將 Net 和 Rank 變回普通列
df_wide = df_wide.reset_index()

print("\n轉換後的寬格式數據：")
print(df_wide)

df_wide = pd.DataFrame()

for value_col in ['Mean', 'Std Dev']:

pivot = df.pivot(index=['Net', 'Rank'], columns='Test', values=value_col)

pivot.columns = [f"{test} {value_col}" for test in pivot.columns]

if df_wide.empty:
    df_wide = pivot
else:
    df_wide = df_wide.join(pivot)

df_wide = df_wide.reset_index()